Artificial Intelligence in Architecture: Applications, Behavioral Analysis, Performance Evaluation, and Challenges in Supporting Architectural Design Practice
Artificial Intelligence in Architecture: Applications, Behavioral Analysis, Performance Evaluation, and Challenges in Supporting Architectural Design Practice
Ali Rakhshani1 Arman Adeli Sabbagh2
1) M.A. Student in Architecture, Faculty of Architecture, Hakim Sabzevari University, Sabzevar, Iran Email:
2) M.A. Student in Architecture, Faculty of Architecture, Hakim Sabzevari University, Sabzevar, Iran Email:
IOI Code
:
AAHF-ABAADH
Publication :
National Conference on Architecture and Art Technologies, Smart City, Soft and Creative Industries and Technological Culture(icact.ir)
Abstract :
The integration of artificial intelligence (AI) into architectural practice has transformed design workflows, visualization processes, and creative exploration. This paper examines the application of generative AI tools—including Midjourney, Stable Diffusion, DALL-E, and large language models such as ChatGPT—in supporting architects across multiple domains: conceptual design, rendering optimization, parametric modeling, technical translation, and educational contexts. Through a comprehensive review of recent empirical studies (2023-2025), this research provides a comparative analysis of AI tool performance, explores prompt engineering techniques tailored to architectural visualization, and investigates behavioral patterns of architect-AI interaction. The study identifies key challenges including dataset biases, functional understanding limitations, ethical concerns regarding authorship, and integration barriers. Quantitative data from adoption studies, controlled experiments, and case analyses reveal that while AI tools significantly accelerate ideation phases and enhance visual exploration, they require careful human oversight to ensure functional correctness and cultural sensitivity. The paper concludes with recommendations for hybrid workflows that combine high-quality concept generations with controllable fine-tuning, emphasizing the need for standardized benchmarks and human-centered evaluation frameworks to guide evidence-based tool selection in architectural practice.
Keywords : artificial intelligence architectural design generative ai prompt engineering design visualization